Survival, Look-Ahead Bias and the Persistence in Hedge Fund Performance

Hedge funds databases are typically subject to high attrition ratesbecause of fund termination and self-selection. Even when all fundsare included up to their last available return, one cannot preventthat ex post conditioning biases a.ect standard estimates ofperformance persistence. In this paper we analyze the persistence inthe performance of U.S. hedge funds taking into account look-aheadbias (multi-period sampling bias). To do so, we model attrition ofhedge funds and analyze how it depends upon historical performance.Next, we use a weighting procedure that eliminates look-ahead bias inmeasures for performance persistence. The results show that the impactof look-ahead bias is quite severe, even though positive and negativesurvival-related biases are sometimes suggested to cancel out. Athorizons of one and four quarters, we find clear evidence of positivepersistence in hedge fund returns, also after correcting forinvestment style. At the two-year horizon, past winning funds tend toperform poorly in the future.survival;performance measurement;investments;individual profiles;hedge funds